What inspired us to build this project was the amount of diagnosis that occurs every day because symptoms are only considered from the perspective of one doctor and not a multitude whose differing opinions could actually lead to a correct diagnosis. Keeping this in mind, we created Health Matters, a self-diagnosing aid, to take in your symptoms and data points such as age and gender and give you an accurate portrayal of what a person at said age and of said gender should look out for. We hope that with this project we are able to bridge the gap between the patient and his or her accurate diagnosis and lead to better preventative medicine.
Our web app allows users to obtain personalized recommendations about clinical preventative measures tailored to his/her age, sex, and/or pregnancy status.
We utilized the healthfinder.gov API and integrated outside datasets for machine learning analysis and visualization. To present these findings, we built a website using HTML/CSS.
We ran into multiple challenges halfway through our project such as finding a decent API which matched all our requirements and was still available for use. We also had trouble connecting the back end of our project to the front end of the project because none of us were well versed with javascript or CGI which put our project on hold for a few hours while we figured out what to do. We also ran into merge conflicts on git because of which we had to stop and evaluate whose requests should be accepted or denied where.
We are proud that we have a fully functional website that conveys all our information and that we able to successfully apply machine learning and data visualization to two major datasets. We are also extremely proud of the fact that a user can get information on demand about what he or she needs to look out for at that stage in their life.
How to use JavaScript, APIs, and how to actually implement machine learning in python
Next steps involve an implementation of a database for providing care and pharmacies related to the personalized results and more characterizations of preventative diseases. We also hope to do a similar analysis for multiple diseases in the future.